The goal to mimic human perception and learning is difficult to achieve, especially in complex environments. As a result of this, a large gap exists between customer needs or unsolved problems and the available technological options, indicating the enormous market potential for Vision+. The task of Vision+ is to fill this technical gap through the use of additional information such as multimodal sensing, spatial-temporal analysis, but also human interactions.
The use of imaging sensors in the industry
Today, imaging sensors are used widely in both industry and everyday life. Applications include visual inspection in industrial manufacturing companies, personal recognition systems, as well as simple portable image and video recording devices, such as those found in mobile phones and cars. To process the ever-growing amounts of data automated analysis is required. Despite progress in this area, many of the image processing algorithms that have been developed are not generally applicable and often not robust enough for complex application areas.
Although the extension of visual systems using human-in-the-loop models (e.g., provision of training data) reduces autonomy, it results in significantly increased robustness. In addition, high-level knowledge is integrated into the training process. Research on modular architectures for image processing systems guarantees the necessary flexibility during the use of modern hardware platforms (e.g., GPGPU, multi-core or mobile computing).
Examination of the results
Ultimately, the Vision+ results will be examined for their applicability in real world scenarios and with respect to their sustainability, and whether they have influence on standardization efforts. The Vision+ consortium demonstrates optimal expertise among academic and applied research groups. Industrial partners who bring great powers of innovation to the application round out the consortium. Vision+ will provide many solutions to previously unresolved challenges. The expected results will mean both a big step towards zero-defect production and significantly higher detection rates for object recognition and classification. In this way, completely new markets can be accessed through wider application fields.
See also: Vision+ Homepage